A remarkable new species of Pterocarpus (Leguminosae: Papilionoideae: Dalbergieae) from Bahia, Brazil

A remarkable new species of Pterocarpus (Leguminosae: Papilionoideae: Dalbergieae) from Bahia, Brazil

Klitgaard, B B

Lewis (1987) cited two collections under Pterocarpus ternatus Rizzini: Lima 13138, the type, and Sarmento 606/80. Additional material shows that these two collections belong to two distinct species. Sarmento 606/80 (a flowering collection) in fact represents a remarkable new simple-leaved species which was apparently first collected in fruit by Davidse & D’Arcy in 1976. The connection between this fruiting specimen and Sarmento 606/80 was overlooked until Queiroz, in 1997, collected more fruiting material which led to a detailed analysis of the species by the three authors. The new species, described below, is restricted to the sand dunes along the margins of the Rio Sao Francisco in central Bahia, Brazil.

Pterocarpus monophyllus Klitgaard, L. P Queiroz & G. P Lewis sp. nov., a speciebus ceteris omnibus generis foliis simplicibus atque fructibus suborbicularibus paene exalatis differens. Typus: Brazil, Bahia, Barra: Ibiraba (=Icatu), Santa Cruz, Queiroz 4825 (holotypus HUEFS, isotypus K).

Shrub to small tree, 1.5 – 4 m tall, crown densely branched, trunk tortuose, bark smooth, dark grey, bark of branchlets dark grey to almost black, without lenticels, present year’s growth arising from short knobbly branchlet apices and lateral brachyblasts. Stipules persistent (becoming lignified) or not, lanceolate-triangular, c. 1.5 x 1 mm. Leaves simple (not unifoliolate); petiole 1.5 – 2 mm long; leaf blade 1.6 – 5.2 x 0.9 – 2.9 cm, elliptic to narrowly or broadly ovate, apex retuse, base rounded to broadly cuneate to subcordate, chartaceous to semi-coriaceous, glabrous and shiny above, very finely appressed-pubescent below, especially on the primary vein (the hairs only apparent at x 10), glabrescent; on dried herbarium material the primary and secondary veins prominent on both surfaces, the tertiary venation prominent above, less so below, the secondary venation brochidodromous, 4 – 6 of the secondaries converging at the blade base, tertiaries reticulate, the lamellar tissue between the veins packed with tannins. Inflorescence a subumbelliform fascicle at branchlet apices or brachyblast tips; peduncles, pedicels, bracteoles, and calyces densely appressed- to very slightly spreading-pubescent; peduncles c. 6 mm long, pedicels c. 3 mm long, bracteoles filiform, margins somewhat toothed, 2 x 0.25 – 0.3 mm. Flowers yellow; calyx tube c. 7 mm long, packed with tannins, teeth 2.5 – 3.5 mm long, margins white-tomentose, standard petal orbicular, apex emarginate, 17 (including a 3 mm claw) x 17 – 18 mm, wings 17 (including a 5 mm claw) x 9 mm, keel petals 12 – 15 (including a 5 mm claw) x 5 – 6 mm, the two petals joined for c. 5 mm along lower margin near apex; petal claws and blade bases packed with tannins round the veins; androecium monadelphous, fused portion of staminal sheath 7 mm long, 9 stamens with filaments free for 3 – 8 mm (measuring from margins to centre of opened out androecium), vexillary stamen with filament free almost to the base, 7.5 mm long, anthers 0.7 x 0.4 mm; ovary densely whitepubescent, style c. 15 mm long, glabrous except for a few hairs at the base, tapering towards a simple terminal, glabrous stigma. Fruit a woody, indehiscent, 1-seeded pod, narrowly winged along half its circumference, 26 – 35 x 26 – 32 x 11 – 16 mm, finely white appressed puberulous, smooth (to naked eye), finely pitted (x 10), yellowish ochre in colour, tannin deposits prolific amongst the lignified venation of the mesocarp. Seed 22 x 11 x 12 mm, smooth, orange-brown, hilum on a slightly raised elliptic boss. Fig. 1.

BRAZIL. Bahia, Mun. Ibiraba, dunas arenosas A margem do Rio Sao Francisco, without date (fl.), Rocha 24 (K, SP); entre Lagoa do Padre e Brejo do Sal, 42’54’W, I O’S, 11 Aug. 1980 (fl.), Sarmento 606/80 (ALCB, HRB, RB, photo K); c. 4 – 5 km N of Xique-Xique on the west side of Rio Sao Francisco, 5 April 1976 (fr.), Davidse &DArcy 11989 (SP); Barra: Ibiraba (=Icatu), Santa Cruz, em frente ao povoado de Ibiraba, no caminho para os Brejos, 42’50’W, 10’48’S, 24 Feb. 1997 (fr.), Queiroz 4825 (holotype HUEFS, isotype K).

HABITAT. Sandy dunes with low forest and cacti along the margins of the Rio Sao Francisco, caatinga, shrubby-arboreal scrub on sand dunes, c. 35 m alt. VERNACULAR NAME. “Capote” (Queiroz 4825).

Rojo (1972) revised Pterocarpus across its entire pantropical range. He recognised only twenty species (six of which occur in the neotropics) after reducing many names to synonymy. Lewis (1987) suggested that a total of 25 – 30 species was probably a more realistic figure, pointing out that Rojo’s lumping of several names under P. rohrii Vahl seemed unsatisfactory and that the genus required further careful study, especially in Brazil.

The combination of simple leaves (these lack a terminal pulvinus on the petiole so that the leaf blade cannot articulate at its point of attachment to the petiole, in contrast to unifoliolate leaves which have a terminal pulvinus and articulated blades) and suborbicular, almost wingless fruits in P monophyllus readily distinguish it from all other species of Pterocarpus. Unifoliolate (not truly simple) leaves occur rarely in depauperate forms of the African taxon P lucens Lepr. subsp. lucens, but normal forms of this and all forms of all other species of the genus are imparipinnate. Most Pterocarpus species have winged fruits with a central seed chamber and are adapted to dispersal by wind. A few species have fruits with highly reduced wings and a thickened, fibrous mesocarp and these are dispersed by water.

In South America P amazonum (Benth.) Amshoff, P santalnoides L’Her. ex DC. and P officinalis Jacq. subsp. officinalis have water-dispersed fruits. P santalinoides has an amphiatlantic distribution while a second subspecies of P officinalis, subsp. gilletii (De Wild.) Rojo, occurs in Congo (Kinshasa) and Congo (Brazzaville), and this also has water-dispersed fruits. All these water dispersed taxa have fruits that are not that unlike those of P monophyllus, although all are generally larger, more laterally compressed and never characteristically suborbicular as in P monophyllus. Preliminary field observations (LPG show that the fruits of P monophyllus simply drop at maturity and remain on the sand below the mother tree. We can speculate that in times of heavy rain or flood the fruits are then dispersed by water, although there are no field data to support this hypothesis. Simple laboratory experiments (GPL) demonstrated that mature fruits have a water-tight seed chamber and float on their sides.

There is no available phylogenetic account for the genus Pterocarpus, and it is presently unknown whether the presumed water-dispersed fruits of P monophyllus reflect a relationship with the similarly dispersed Amazonian species, or if they have evolved independently within an eastern Brazilian group of species. To date P monophyllus is known from very few collections, all of which were made close to the town of Ibiraba. The total area covered by the dune systems of the Sao Francisco valley is about 7000 km2 but, based on available collections, P monophyllus appears to be a narrowly restricted endemic and we cannot assume that the species is widespread elsewhere in the dunes. Nevertheless, such restricted endemism would be rare for a species of the arboreal caatinga. Given the large expanse of dunes available for colonisation it seems unlikely that the restricted distribution of the species is relictual, but rather that it is due either to poor dispersal ability or to under-collecting in the area, or both. This is, however, pure speculation. For most of the year the dunes lie approximately one kilometre from the margin of the Rio Sao Francisco, but after heavy rains when the river is in flood the varzea zone between the river margin and the dunes becomes inundated, thus bringing the river and the dunes into direct contact. Furthermore, the species does not occur just along the limits of the dune system but is also found a little way into their interior, so that the species cannot be considered to be of riverine habitat.

AcKNOWLEDGEMENTS

LPQ thanks Pedro Luis Rocha who provided field work support, and all three authors thank Alan Radcliffe-Smith for the Latin diagnosis, Peter Gasson and Hazel Wilkinson for confirming the presence of tannins in the leaves and fruits, Haroldo Cavalcante de Lima for sending photographs of Pterocarpus ternatus, Camilla Speight for the beautiful illustration, R. Toby Pennington for a constructive review, and the curators of herbaria cited who loaned material for study.

Copyright The Controller of Her Majesty’s Stationery Office 2000

Provided by ProQuest Information and Learning Company. All rights Reserved.